This is a Monte Carlo based level III fugacity program for the 5 Great Lakes. It treats the half-lives in the relevant environmental media and certain physical properties of the chemical as probabilistic input parameters. It is based on a four compartment level III multi-media fugacity program (1), which has been parameterized specifically for each drainage basin. The program also runs a standard level III fugacity model for each lake, using single point input variables. It is intended as a screening level tool and is not recommended for regulatory decision-making.
Frequently Asked Questions
Q: What does the model output and how should the results be interpreted?
A: Like all Level III programs it should be understood that steady-state conditions apply. This is a very reasonable assumption for most of the major pollutants which enter the Great Lakes ecosystem. Pulse loadings from a one time event are not modeled. Outputs include the estimated environmental concentrations (EEC) that can be expected in the environmental media and the overall persistence time of the chemical in the basin. If the estimation programs are used various aquatic toxicity endpoints are also estimated.
Q: What inputs are required?
A: The user must specify the following properties for the chemical: Henry’s Law constant, the octanol/water partition coefficient (log Kow), melting point, vapor pressure, molecular weight, and the half-lives in each environmental medium. If using the Monte Carlo mode the half-lives must be given as either lognormal or triangular distributions. The coefficient of variation (CV) for the Henry’s Law constant, log Kow, melting point, and vapor pressure are fixed within the model. Emissions to one of the compartments must also be included.
Q: What if I don’t know a value for some of the inputs because no values are available or this is a new chemical which does not have measured values?
A: One of the strengths of this program is that all of the properties of the chemical can be estimated for the user. Simply input the SMILES notation of the chemical and all the properties will be estimated by the program.
Q: How do I create a SMILES notation?
A: Most commercially available chemical drawing software programs allow you to draw a structure and get its SMILES notation. You can also use the drawing program we have incorporated into the model or an appropriate CAS registry number and then submit the SMILES notation to the program.
Q: How do I estimate the emission rate into the ecosystem?
A: This is a more difficult problem since direct emissions into the lake itself or the air surrounding it are usually not known. One source of emission data is from the Toxic Release Inventory maintained by the EPA. Another source of air emissions to the Great Lakes region is the Great Lakes Regional Air Toxics Inventory Report from the GLC. Direct emissions into water can be estimated by using monitoring and hydrological data from waterways that flow directly into the Lake. For example, the majority of water that enters Lake Ontario arrives from the Niagara River via Lake Erie. The water flow rate of the Niagara River into Lake Ontario is approximately 6,000 m3/second or approximately 1.89X1011 m3/year. Knowing the concentration of the chemical in the river and multiplying it by the flow rate will give an estimate of the loading rate into the lake. We have also created Pesticide Use Maps for the Great Lakes States and maps for the 2002 Great Lakes Toxic Air Emissions Inventory from the Great Lakes Comission. More data are available at: CAROL
Q: What ecotoxicological endpoints can be estimated?
A: The estimation program employs the ECOSAR model which uses structure-activity relationships (SARs) to predict the aquatic toxicity of chemicals based on their similarity of structure to chemicals for which the aquatic toxicity has been previously measured. For certain classes of chemicals, modeled endpoints include LC50 estimates for fish, mysid shrimp, daphnid, and earthworms; EC50 values for daphnid and algae; chronic toxicity estimates for fish and algae.
Q: How can I use the model to develop an emission reduction strategy?
A: The model contains a stochastic linear searching algorithm that employs the user specified input parameters of the model and the initial set of emissions as an initial guess and estimates an emission rate required to drop the average concentarion of the chemical below a user defined level of concern (LOC) in the Lake.
Q: What do the outputs mean?
The outputs tell you where a chemical is expected to partition in the ecosystem and how long it is expected to stay in this specific environment (Persistence Time). If run in Monte Carlo mode, you obtain a distribution or percentile of the results. For instance if the 5th percentile of the persistence time is 100 hours, there is a 5% probability that the persistence time is 100 hours or less, if the 50th percentile is 200 hours, there is a 50% probability that the persistence time is 200 hours or less, if the 95th percentile is 400 hours, there is a 95% probability that the persistence time is 400 hours or less, and so forth.
Q: Does the persistence time mean that the entire chemical has degraded?
A: No, it only tells you how long it has remained in this specific environment, as advection processes such as wind and water flow may have carried it out of the ecosystem without being degraded.
Q: How was the model parameterized for the individual basins?
A: Meteorological and hydrological data was obtained for the area and used to refine the advection rates, volumes, and mass transport parameters of the model. This data was then used to develop the model and several well studied chemical contaminants were used to help validate the model.
Q: Is sensitivity analysis performed?
A: Part of the output includes the Pearson rank correlation coefficient . This coefficient gives an indication of the importance of a particular input variable to a specific output parameter. The closer the coefficient is to either -1 or 1, the stronger the linear correlation between the variables. A value near zero suggests little correlation.
Q: What is the characteristic travel distance (CTD) and long range transport (LRT)?
A: These quantities are conceptual properties that provide an indication of whether a chemical released to air can be transported undegraded over long distances. The CTD is calculated using a generic level III fugacity model that does not account for advective processes and the LRT potential is ranked according to a previously published ranking scheme(2). The absolute value of the CTD should not be taken literally since it will vary depending upon which multi-media fate model is used to calculate it and the environmental properties of the model; however, the trends for various chemicals should be similar, which allows them to be ranked as having either low, moderate or high LRT potential.
(1) Citra MJ. Incorporating Monte Carlo Analysis into Multi-Media Environmental Fate Models Environ Toxicol Chem 23: 1629-1633 (2004).
(2) Beyer A, Mackay D, Matthies M, Wania F, Webster E. Assessing Long-Range Transport Potential of Persistent Organic Chemicals Environ Sci Technol 34: 699-703 (2000).